This application responds to the Funding Opportunity Announcement PAR 06-248 "From Intervention Development to Services: Exploratory Research Grants" (R34), and addresses a longstanding NIMH priority of improving medication adherence for persons with schizophrenia. Despite other treatment advances, nonadherence remains intractable. Nowhere is this more true than for patients who are recovering from a "first- episode" schizophrenia, who tend to respond very well to medication but almost always stop taking their antipsychotic prematurely, leading to devastating consequences. Improving treatment adherence is perhaps the most promising strategy for improving clinical outcomes in schizophrenia. We have developed a new intervention for first episode patients that is based on a Cognitive Behavior Therapy (CBT) platform. Adherence interventions for other conditions can at least assume that their patient population has accepted their diagnosis. For persons with schizophrenia, this basic assumption often does not hold. Patient rejection of a disease model severely limits the usefulness of adherence interventions that rely on such a model. We propose to finalize and then test an alternate strategy that addresses adherence issues indirectly, by working with rather than against the patient's current belief system. The PI and his collaborators are developing a new intervention known as Health Dialogue Intervention (HDI) based on a CBT platform. The therapeutic focus of HDI is to work with the patient's inherent goals to achieving desired health outcomes using their own perspective, rather than insisting on the clinicians'disease model. An initial version of HDI has shown encouraging results from a small randomized pilot study comparing HDI vs. Treatment as Usual (TAU) in recently stabilized, multi-episode schizophrenia patients. The first goal of this R34 is to further refine and finalize the HDI intervention, using additional material from our adherence research, and from patient narrative research. The second goal is to test the HDI in a more rigorous pilot study, now with recently stabilized "first-episode" instead of multiepisode patients, and against an "active" psychosocial comparator rather than TAU. We will randomize 40 recent onset, stabilized schizophrenia subjects to receive either HDI or psychoeducation. Both interventions will be delivered flexibly with up to 16 individual sessions over the course of 52 weeks. Outcomes will include: 1) acceptance and attendance rates of therapy sessions, 2) effects on adherence attitudes, and 3) effects on adherence behavior. Should there be an efficacy signal, this R34 project will provide preliminary data and guidance for designing a fully powered RCT using HDI in a future R01 application. We believe that the HDI is a very promising and even potentially breakthrough new approach to the vexing and unsolved public health problem of treatment nonadherence in schizophrenia. Schizophrenia can be a devastating illness, but better outcomes are possible. Continued treatment with antipsychotic medication is necessary to achieve better outcomes, but nonadherence (stopping medications too soon) remains a serious and unresolved problem. This application has 2 main research goals that hopefully will result in new approaches to this problem. The first goal is to finalize an innovative, patient-centered therapy that we believe will help patients stay on their medications, even when the patient does not agree with the doctors about their diagnosis of schizophrenia. Second, we plan to test this treatment in a group of recently stabilized, "first-episode" persons diagnosed with schizophrenia who we know from other research studies will be very likely to stop their medications too soon. We will compare what happens in between groups of patients randomized to this treatment with another, more standard psychoeducation treatment, in terms of differences in attitudes towards medications, how long patients remain on medication, and their willingness to resume medications after stopping and symptoms return.